Multi-Scale Computation


Computational mechanics seeks to develop new, computer-aided methods for predicting physical phenomenon important to engineering. We leverage the resources of the parallel computing cluster maintained by Advanced Research Computing at U-M to perform large scale computations.

We use multi-scale computational methods to research questions ranging from the molecular basis of soot formation in combustion to the manner in which molecular-level defects affect macroscopic mechanical properties. These methods focus on predicting the mechanical, electrical, and optical behavior of materials and structures from smaller scale models in an accurate and reliable way. This can also involve quantum-mechanical calculations or complex substructure models.


  • Simulation of turbulence
  • Structural health monitoring and biodynamics
  • Biomechanics and electroacoustics
  • Phononic material design and computational mechanics
  • Combustion and reacting flows
  • Computational fluid dynamics
  • Optimization and homogenization methods
  • DNA mechanics and dynamics
  • Computational physics
  • Computational materials physics

Recent News

The 2020 Leadership in Engineering Award from the Washington Academy of Sciences has been awarded to Dawn Tilbury.

This award is presented annually to senior graduate students who have demonstrated excellence in research and scholarship in an area of applied mechanics.

Research from the Computational Materials Physics group, directed by Associate Professor Vikram Gavini, selected as a finalist for the ACM Gordon Bell Prize.

Krishna Garikipati receives a fellowship for work in numerical methods applied to nonlinear problems.

Two awardees have received the 2019 Clare Boothe Luce Fellowships for a PhD in Scientific Computing program.

An award for outstanding achievement in the mechanical engineering field.